Inhomogeneous Dependence Modeling with Time-Varying Copulae
使用时变参数的Copula函数建模金融收益序列的非正态依赖关系，并在投资组合风险价值估计中优于RiskMetrics方法。
Measuring dependence in multivariate time series is tantamount to modeling its dynamic structure in space and time. In risk management, the nonnormal behavior of most financial time series calls for non-Gaussian dependences. The correct modeling of non-Gaussian dependences is, therefore, a key issue in the analysis of multivariate time series. In this article we use copula functions with adaptively estimated time-varying parameters for modeling the distribution of returns. Furthermore, we apply copulae to the estimation of Value-at-Risk of portfolios and show their better performance over the RiskMetrics approach.